798 SUBANTARCTIC ISLANDS OF NEW ZEALAND. [Summary of Results. 



However striking these resemblances may appear, it must be remembered 

 that nearly all these plants have seeds adapted for wide dispersal by birds, wind, or 

 ocean-currents, and attention must therefore be directed to the terrestrial animals, 

 which are much less likely to be carried across the ocean by means of this kind. 



The fresh-water fish found on the Auckland and Campbell Islands belong to the 

 Galaxiidae, a family that has representatives in New Zealand, South America, Falk- 

 land Islands, Australia, and South Africa, one species {Galaxias attenuatus) having 

 a very wide range, being found in New Zealand, Tasmania, southern Australia, Falk- 

 land Islands, and the extremity of South America. This species is known to breed 

 in the sea, and it has been supposed that this accounts for its wide distribution ; 

 and as another species {G. hollansi) was originally described as a marine one Dr. G. A. 

 Boulenger (1902) has used this as corroborative evidence, and appears to consider 

 that the Galaxiidae can no longer be used as evidence of former land connection. 

 A reference to Mr. Waite's paper (p. 586) shows that the description of G. hollansi 

 as marine is probably erroneous, and, as Mr. Geoffrey Smith (1909a, p. 138) has 

 pointed out, the fact that species of Galaxiidae breed in the sea by no means does 

 away with all the value of the group in favour of land connection, or proves that 

 they can readily cross the wide oceans. 



Among the spiders Mr. Hogg (p. 156) calls attention to the preponderating 

 number of specimens, as well as to the area of distribution of the genera Myro and 

 Rubrius, and points out that this group stands out as quite an antarctic type. Of 

 the genus Myro, one species is found at Macquarie Island, two others at the Snares, 

 a fourth at Kerguelen Island, and another at the Cape of Good Hope ; of Rubrius, 

 Mr. Hogg describes three species, one of which {R. falxiatus) is said to approach 

 more nearly to a South American species than to the Tasmanian forms of the 

 genus. The species Pacificana cockayni, from Bounty Island, though referred to a 

 genus of its own, shows affinities to other genera recorded from Cape Horn and from 

 Tasmania. 



Among the earthworms, Notiodrilus (s.l.) is a genus found in New Zealand and 

 its outlying islands, in Kerguelen, Marion Island, Crozets, South Georgia, Falkland 

 Islands, Patagonia, and the Cape of Good Hope, and is therefore distinctly circum- 

 austral, though some species extend to Mexico, Madagascar, north-west Aus- 

 tralia, &c. The genus Phreodrilus is also characteristically subantarctic, and the 

 family Haplotaxidae is largely represented in the Southern Hemisphere. The new 

 species of Lumbricillus is closely related to two other species which occur in sub- 

 antarctic regions, though the genus is found all over the Northern Hemisphere. 

 The genera Marionina and Enchytraeus, though of northern origin, have species in 

 subantarctic regions. The subfamily Megascolecinae is Australian in origin, but 

 has representatives in the North Island of New Zealand, while it is represented in 

 the New Zealand subantarctic islands by four species of Diporochaeta and one species 

 of Plutellus, a distinctly Australian genus ; Professor Benham, however, is inclined 

 to think that this species (P. aucklandicus) may have been introduced into the Auck- 

 land Islands by man. 



The earthworms thus appear to show close connection with those of other sub- 

 antarctic lands, and it has been supposed by Beddard and Benham that this could 

 only be accounted for by the existence of former land connection, such as might 

 be supplied by an antarctic continent. Dr. Michaelsen (1902, p. 158) has pointed 



